X-Git-Url: https://git.saurik.com/redis.git/blobdiff_plain/ebd85e9a455df689c9be02a93354f580df4cafd8..89423052cab4700291f290819d53f4fbb8dabf95:/src/rdb.c diff --git a/src/rdb.c b/src/rdb.c index c74ec41c..5018473e 100644 --- a/src/rdb.c +++ b/src/rdb.c @@ -1,6 +1,7 @@ #include "redis.h" #include "lzf.h" /* LZF compression library */ #include "zipmap.h" +#include "endianconv.h" #include #include @@ -20,17 +21,15 @@ int rdbSaveType(rio *rdb, unsigned char type) { return rdbWriteRaw(rdb,&type,1); } +/* Load a "type" in RDB format, that is a one byte unsigned integer. + * This function is not only used to load object types, but also special + * "types" like the end-of-file type, the EXPIRE type, and so forth. */ int rdbLoadType(rio *rdb) { unsigned char type; if (rioRead(rdb,&type,1) == 0) return -1; return type; } -int rdbSaveTime(rio *rdb, time_t t) { - int32_t t32 = (int32_t) t; - return rdbWriteRaw(rdb,&t32,4); -} - time_t rdbLoadTime(rio *rdb) { int32_t t32; if (rioRead(rdb,&t32,4) == 0) return -1; @@ -437,7 +436,8 @@ int rdbSaveObjectType(rio *rdb, robj *o) { return -1; /* avoid warning */ } -/* Load object type. Return -1 when the byte doesn't contain an object type. */ +/* Use rdbLoadType() to load a TYPE in RDB format, but returns -1 if the + * type is not specifically a valid Object Type. */ int rdbLoadObjectType(rio *rdb) { int type; if ((type = rdbLoadType(rdb)) == -1) return -1; @@ -602,10 +602,12 @@ int rdbSave(char *filename) { dictIterator *di = NULL; dictEntry *de; char tmpfile[256]; + char magic[10]; int j; long long now = mstime(); FILE *fp; rio rdb; + uint64_t cksum; snprintf(tmpfile,256,"temp-%d.rdb", (int) getpid()); fp = fopen(tmpfile,"w"); @@ -616,7 +618,10 @@ int rdbSave(char *filename) { } rioInitWithFile(&rdb,fp); - if (rdbWriteRaw(&rdb,"REDIS0003",9) == -1) goto werr; + if (server.rdb_checksum) + rdb.update_cksum = rioGenericUpdateChecksum; + snprintf(magic,sizeof(magic),"REDIS%04d",REDIS_RDB_VERSION); + if (rdbWriteRaw(&rdb,magic,9) == -1) goto werr; for (j = 0; j < server.dbnum; j++) { redisDb *db = server.db+j; @@ -644,9 +649,17 @@ int rdbSave(char *filename) { } dictReleaseIterator(di); } + di = NULL; /* So that we don't release it again on error. */ + /* EOF opcode */ if (rdbSaveType(&rdb,REDIS_RDB_OPCODE_EOF) == -1) goto werr; + /* CRC64 checksum. It will be zero if checksum computation is disabled, the + * loading code skips the check in this case. */ + cksum = rdb.cksum; + memrev64ifbe(&cksum); + rioWrite(&rdb,&cksum,8); + /* Make sure data will not remain on the OS's output buffers */ fflush(fp); fsync(fileno(fp)); @@ -662,6 +675,7 @@ int rdbSave(char *filename) { redisLog(REDIS_NOTICE,"DB saved on disk"); server.dirty = 0; server.lastsave = time(NULL); + server.lastbgsave_status = REDIS_OK; return REDIS_OK; werr: @@ -688,7 +702,7 @@ int rdbSaveBackground(char *filename) { if (server.ipfd > 0) close(server.ipfd); if (server.sofd > 0) close(server.sofd); retval = rdbSave(filename); - _exit((retval == REDIS_OK) ? 0 : 1); + exitFromChild((retval == REDIS_OK) ? 0 : 1); } else { /* Parent */ server.stat_fork_time = ustime()-start; @@ -698,6 +712,7 @@ int rdbSaveBackground(char *filename) { return REDIS_ERR; } redisLog(REDIS_NOTICE,"Background saving started by pid %d",childpid); + server.rdb_save_time_start = time(NULL); server.rdb_child_pid = childpid; updateDictResizePolicy(); return REDIS_OK; @@ -719,7 +734,7 @@ robj *rdbLoadObject(int rdbtype, rio *rdb) { size_t len; unsigned int i; - redisLog(REDIS_DEBUG,"LOADING OBJECT %d (at %d)\n",rdbtype,rdb->tell(rdb)); + redisLog(REDIS_DEBUG,"LOADING OBJECT %d (at %d)\n",rdbtype,rioTell(rdb)); if (rdbtype == REDIS_RDB_TYPE_STRING) { /* Read string value */ if ((o = rdbLoadEncodedStringObject(rdb)) == NULL) return NULL; @@ -788,7 +803,7 @@ robj *rdbLoadObject(int rdbtype, rio *rdb) { } /* This will also be called when the set was just converted - * to regular hashtable encoded set */ + * to a regular hash table encoded set */ if (o->encoding == REDIS_ENCODING_HT) { dictAdd((dict*)o->ptr,ele,NULL); } else { @@ -843,9 +858,10 @@ robj *rdbLoadObject(int rdbtype, rio *rdb) { hashTypeConvert(o, REDIS_ENCODING_HT); /* Load every field and value into the ziplist */ - while (o->encoding == REDIS_ENCODING_ZIPLIST && len-- > 0) { + while (o->encoding == REDIS_ENCODING_ZIPLIST && len > 0) { robj *field, *value; + len--; /* Load raw strings */ field = rdbLoadStringObject(rdb); if (field == NULL) return NULL; @@ -854,23 +870,27 @@ robj *rdbLoadObject(int rdbtype, rio *rdb) { if (value == NULL) return NULL; redisAssert(field->encoding == REDIS_ENCODING_RAW); + /* Add pair to ziplist */ + o->ptr = ziplistPush(o->ptr, field->ptr, sdslen(field->ptr), ZIPLIST_TAIL); + o->ptr = ziplistPush(o->ptr, value->ptr, sdslen(value->ptr), ZIPLIST_TAIL); /* Convert to hash table if size threshold is exceeded */ if (sdslen(field->ptr) > server.hash_max_ziplist_value || sdslen(value->ptr) > server.hash_max_ziplist_value) { + decrRefCount(field); + decrRefCount(value); hashTypeConvert(o, REDIS_ENCODING_HT); break; } - - /* Add pair to ziplist */ - o->ptr = ziplistPush(o->ptr, field->ptr, sdslen(field->ptr), ZIPLIST_TAIL); - o->ptr = ziplistPush(o->ptr, value->ptr, sdslen(value->ptr), ZIPLIST_TAIL); + decrRefCount(field); + decrRefCount(value); } /* Load remaining fields and values into the hash table */ - while (o->encoding == REDIS_ENCODING_HT && len-- > 0) { + while (o->encoding == REDIS_ENCODING_HT && len > 0) { robj *field, *value; + len--; /* Load encoded strings */ field = rdbLoadEncodedStringObject(rdb); if (field == NULL) return NULL; @@ -915,12 +935,13 @@ robj *rdbLoadObject(int rdbtype, rio *rdb) { { unsigned char *zl = ziplistNew(); unsigned char *zi = zipmapRewind(o->ptr); + unsigned char *fstr, *vstr; + unsigned int flen, vlen; + unsigned int maxlen = 0; - while (zi != NULL) { - unsigned char *fstr, *vstr; - unsigned int flen, vlen; - - zi = zipmapNext(zi, &fstr, &flen, &vstr, &vlen); + while ((zi = zipmapNext(zi, &fstr, &flen, &vstr, &vlen)) != NULL) { + if (flen > maxlen) maxlen = flen; + if (vlen > maxlen) maxlen = vlen; zl = ziplistPush(zl, fstr, flen, ZIPLIST_TAIL); zl = ziplistPush(zl, vstr, vlen, ZIPLIST_TAIL); } @@ -930,8 +951,11 @@ robj *rdbLoadObject(int rdbtype, rio *rdb) { o->type = REDIS_HASH; o->encoding = REDIS_ENCODING_ZIPLIST; - if (hashTypeLength(o) > server.hash_max_ziplist_entries) + if (hashTypeLength(o) > server.hash_max_ziplist_entries || + maxlen > server.hash_max_ziplist_value) + { hashTypeConvert(o, REDIS_ENCODING_HT); + } } break; case REDIS_RDB_TYPE_LIST_ZIPLIST: @@ -986,6 +1010,8 @@ void startLoading(FILE *fp) { /* Refresh the loading progress info */ void loadingProgress(off_t pos) { server.loading_loaded_bytes = pos; + if (server.stat_peak_memory < zmalloc_used_memory()) + server.stat_peak_memory = zmalloc_used_memory(); } /* Loading finished */ @@ -1009,6 +1035,8 @@ int rdbLoad(char *filename) { return REDIS_ERR; } rioInitWithFile(&rdb,fp); + if (server.rdb_checksum) + rdb.update_cksum = rioGenericUpdateChecksum; if (rioRead(&rdb,buf,9) == 0) goto eoferr; buf[9] = '\0'; if (memcmp(buf,"REDIS",5) != 0) { @@ -1018,7 +1046,7 @@ int rdbLoad(char *filename) { return REDIS_ERR; } rdbver = atoi(buf+5); - if (rdbver < 1 || rdbver > 3) { + if (rdbver < 1 || rdbver > REDIS_RDB_VERSION) { fclose(fp); redisLog(REDIS_WARNING,"Can't handle RDB format version %d",rdbver); errno = EINVAL; @@ -1032,7 +1060,7 @@ int rdbLoad(char *filename) { /* Serve the clients from time to time */ if (!(loops++ % 1000)) { - loadingProgress(rdb.tell(&rdb)); + loadingProgress(rioTell(&rdb)); aeProcessEvents(server.el, AE_FILE_EVENTS|AE_DONT_WAIT); } @@ -1071,8 +1099,12 @@ int rdbLoad(char *filename) { if ((key = rdbLoadStringObject(&rdb)) == NULL) goto eoferr; /* Read value */ if ((val = rdbLoadObject(type,&rdb)) == NULL) goto eoferr; - /* Check if the key already expired */ - if (expiretime != -1 && expiretime < now) { + /* Check if the key already expired. This function is used when loading + * an RDB file from disk, either at startup, or when an RDB was + * received from the master. In the latter case, the master is + * responsible for key expiry. If we would expire keys here, the + * snapshot taken by the master may not be reflected on the slave. */ + if (server.masterhost == NULL && expiretime != -1 && expiretime < now) { decrRefCount(key); decrRefCount(val); continue; @@ -1085,6 +1117,20 @@ int rdbLoad(char *filename) { decrRefCount(key); } + /* Verify the checksum if RDB version is >= 5 */ + if (rdbver >= 5 && server.rdb_checksum) { + uint64_t cksum, expected = rdb.cksum; + + if (rioRead(&rdb,&cksum,8) == 0) goto eoferr; + memrev64ifbe(&cksum); + if (cksum == 0) { + redisLog(REDIS_WARNING,"RDB file was saved with checksum disabled: no check performed."); + } else if (cksum != expected) { + redisLog(REDIS_WARNING,"Wrong RDB checksum. Aborting now."); + exit(1); + } + } + fclose(fp); stopLoading(); return REDIS_OK; @@ -1102,14 +1148,19 @@ void backgroundSaveDoneHandler(int exitcode, int bysignal) { "Background saving terminated with success"); server.dirty = server.dirty - server.dirty_before_bgsave; server.lastsave = time(NULL); + server.lastbgsave_status = REDIS_OK; } else if (!bysignal && exitcode != 0) { redisLog(REDIS_WARNING, "Background saving error"); + server.lastbgsave_status = REDIS_ERR; } else { redisLog(REDIS_WARNING, "Background saving terminated by signal %d", bysignal); rdbRemoveTempFile(server.rdb_child_pid); + server.lastbgsave_status = REDIS_ERR; } server.rdb_child_pid = -1; + server.rdb_save_time_last = time(NULL)-server.rdb_save_time_start; + server.rdb_save_time_start = -1; /* Possibly there are slaves waiting for a BGSAVE in order to be served * (the first stage of SYNC is a bulk transfer of dump.rdb) */ updateSlavesWaitingBgsave(exitcode == 0 ? REDIS_OK : REDIS_ERR);